UC Berkeley

Bacteriophages (phages; viruses that infect bacteria) are abundant, diverse, and play important roles in shaping microbial community composition. Phages can structure the composition of bacterial communities by being a direct cause of mortality, altering competitive outcomes, and even providing novel functions to their hosts. Bacteria-phage interactions are highly context-dependent, with environmental variables such as temperature, resource availability, and spatial structure significantly impacting their ecological and (co)evolutionary dynamics. However, while it is known that context is important, relatively little experimental work has explored their interactions in ecologically-relevant conditions. For plant-associated bacteria and phage in particular, there are many open questions about their ecology and evolution in the phyllosphere (aboveground plant tissues). In this dissertation, I explore whether and how context impacts phage interactions with the bacterial plant pathogen Pseudomonas syringae, with a focus on the role of the phyllosphere environment. By leveraging the power of experimental evolution, I specifically tested the impacts of coevolution, the foliar environment, and biofilm formation on interactions between P. syringae and several obligately lytic phages. In Chapter 1, I first explored the genotypic and phenotypic consequences of coevolution between P. syringae and a single lytic phage. I serially passaged this phage with P. syringae under conditions where the bacterium was either allowed to coevolve or was held constant. I then compared the evolved changes in these two groups of phages to determine how coevolution differed from general adaptation to the bacterial host. At the endpoint, individual phages were isolated and genome resequenced to determine identity and frequencies of mutations across populations. Additionally, I quantified the ability of the phages to infect bacterial hosts from their own and alternate populations across time. I found that coevolved phages had more mutations than phages evolved on a constant host, and most mutations were in genes encoding structural proteins. Coevolved phages best reduced growth of past and sympatric hosts, and had a stronger relationship between pairwise genotypic-phenotypic distances than phages evolved on a constant host. In Chapter 2, I next tested the impact of environmental context on the evolution of bacterial resistance to phage. Specifically, I experimentally evolved replicate populations of P. syringae with lytic phages both in vitro (in high-nutrient liquid cultures) and in planta (syringe-infiltrated into the leaf apoplast) and quantified frequencies of phage-resistant colonies over time. Due to a previously described context-dependent cost of phage resistance in planta, we hypothesized that the spread of evolved resistance and coevolutionary dynamics would be slower in the leaf environment. In two separate experimental evolution projects with high amounts of phage, I found that resistant bacteria were virtually undetectable in the plant environment at the end of each experiment. In contrast, resistance reached high frequencies in liquid culture. A follow up assay determined that low rates of phage replication coupled with a slight cost of phage resistance eliminated the fitness benefit of resistance in planta. In Chapter 3, I tested a seedling-based method for screening phage effectiveness as a biocontrol method for P. syringae. In three trials, phages were prophylactically applied to tomato seedlings in sterile conical tubes before flooding with the bacterial pathogen Pseudomonas syringae pathovar tomato DC3000. I recorded seedling disease progression and quantified endpoint bacteria and phage densities in each trial. Phages replicated in all trials, but reduction of disease symptoms and endpoint P. syringae density varied across trials with different application densities. This resource-efficient method rapidly identified an effective phage and application density to mitigate disease on seedlings. In the final chapter (Chapter 4), I tested how biofilm formation impacts phage interactions on both ecological and evolutionary timescales. Biofilms are multidimensional structures composed of cells and an extracellular matrix that can act as a barrier from external stressors such as antibiotics or phages. I hypothesized that biofilm formation may be one factor that contributed to the low rates of phage replication observed in Chapter 2. To test this, I first completed a series of experiments where I grew P. syringae biofilms until a certain time point, applied phage lysates, and then quantified biofilm biomass after phage addition. Across all experiments, phage application never reduced biofilm biomass, and in some cases increased biomass. Next, I experimentally selected for or against P. syringae biofilm formation in the absence or presence of phage selection pressure in a fully factorial design. I identified changes in population and individual-level biofilm formation and phage resistance, and found that phage-resistant isolates invested relatively more in biofilm formation than phage-sensitive isolates. In a seedling inoculation assay with a subset of isolates, phage resistance and high investment in biofilm formation were associated with a cost of virulence. In addition, some phage-resistant mutants caused increased disease symptoms in the presence of one phage. These results offer insight into the traits that might impact phage effectiveness in agricultural biocontrol. Together, this dissertation contributes to a growing body of work exploring context-dependency in bacteria-phage interactions. This work demonstrates that incorporating ecologically-relevant context can yield fundamentally different insights into their ecology and evolution than in vitro research using standard culture conditions.